The field of this invention relates to absorption spectroscopy, more specifically, it pertains to medical uses of absorption spectroscopy to quantify component concentrations in human gas emissions, such as breath and gas emitted through the skin.
Dialysis Population
It is estimated that in the United States, approximately 246,000 patients underwent kidney dialysis treatment in 1999, “Living ESRD Patients on December 31,” Table D.1, United States Renal Data System (USRDS), 2000 ADR/Reference Tables, Section D—Treatment Modalities, www.usrds.org. Most of these patients undergo the treatment in clinics, hospitals, or specialized dialysis centers. However, a sizable number of the patients are able to avail themselves of dialysis through peritoneal dialysis treatments at home, Id. The typical kidney dialysis station costs approximately $25,000. For an effective deployment of such an investment, it is necessary to treat patients as efficiently as possible. It has been observed that ammonia concentration in exhaled human breath of a dialysis patient undergoing hemodialysis drops from over 10,000 ppB (parts per billion) to just over 1,000 ppB during the dialysis. Davies et al., “Quantitative Analysts of Ammonia on the Breath of Patients in End-Stage Renal Failure,” Kidney International 52:223-228 (1997). Physicians often use the smell of a patient's breath as one indicator of health and well-being. For detecting ammonia, this technique is not very sensitive, as Is seen from the fact that the lower limit of human perception for the presence of ammonia through smell is approximately 53 ppm (parts per million), Merck Index, 10th ed., p. 74. Accordingly, the physical examination of a patient by a physician employing simply smelling of patient breath to gather information regarding the status of patient kidney function is impractical, save for the most severe cases.
On the oasis of the above discussion, a reliable and quantitative-measurement of ammonia concentration in breath would be an excellent diagnostic tool for ascertaining incipient kidney trouble, the need for immediate dialysis treatment, determining the efficacy of the procedure during dialysis, and detecting the end-point of dialysis treatment. Determining the scheduling of dialysis treatment through actual measurements of ammonia concentrations in breath would be far superior to providing dialysis at fixed and predetermined intervals and durations. One of the key factors in favor of end-point detection of dialysis treatment is that the patient would not have to remain subjected to the dialysis procedure for any time longer than necessary. From the patient's viewpoint, less time spent connected to a dialysis apparatus equates to reduced physical and perhaps emotional discomfort. From the physician's viewpoint, reliable, accurate, real-time information on the progress of treatments equates to an improved ability to respond to a patient's changing treatment needs. From the viewpoint of the dialysis treatment providers, providing dialysis when called for through an accurate determination of the need for dialysis, and for only the necessary length of time, would allow for more efficient usage of the dialysis facilities and associated medical personnel. Thus, there has long been a need for accurate end-point detection during the dialysis procedure. Quantitative determination of ammonia levels in breath offers a fast, painless solution at a reasonable cost.
Turning now from the situation in a dialysis facility to the home dialysis section of the market, USRDS data indicates that at present approximately $265 million associated with home dialysis were covered by Medicare payments, “Medicare Payments for ESRD Patients,” Table K.1, USRDS 2000 ADR/Reference Tables, Section K—Economic Costs of ESRD, www.usrds.org. For these home health care patients, the ability to non-invasively monitor their kidney health status by measurement of breath ammonia would provide two significant benefits. The first is that the individual could undergo his or her treatment when indicated by elevated levels of breath ammonia, a surrogate for elevated blood urea nitrogen. The second benefit is that the measurements of breath ammonia made during the dialysis treatment would provide an accurate end-point for the treatment. Such in-home dialysis patients could obtain the benefits of such technology in the absence of a trained health care professional.
The total Medicare payments for ESRD patients in 1998 amounted to approximately $11 billion, “Medicare Payments for ESRD Patients,” Table K.1, USRDS 2000 ADR/Reference Tables, Section K—Economic Costs of ESRD, www.usrds.org. A large fraction of these costs are attributable to present treatment methods which rely on regularly scheduled treatments for prescribed lengths of time. Changing to a treatment protocol based on objective measures of treatment efficacy and efficiency will have a major impact on a very large and growing cost base. In addition to being an indicator for the need for dialysis treatment, the presence of ammonia in a patient's breath is also expected to be an indicator of liver transplant success, kidney and fiver function in premature babies, and an indicator for preeclampsia in women during late stage pregnancies.
Asthmatic Population
The number of asthmatic individuals in the United States has been estimated between 14.6 million and 17.2 million patients, “Vital and Health Statistics,” Current Estimates From the National Health Interview Survey, 1994 (Series 10: Data for the National Health Survey No. 193, DHHS Publication No. 96-1521), p. 94. Of these, approximately 10.4 million patients are classified as suffering from chronic asthma. The 1987 National Medical Expenditure Survey results, adjusted to 1996 dollars, show that the direct medical costs associated with asthma patients, including direct hospital outpatient services, hospital inpatient stays, emergency department visits, physician and facility payments and prescribed medicines are in excess of $5 billion, Smith et al., “A National Estimate of Economic Costs of Asthma,” Am. J. Respir. Crit. Care Med, 156, 787-793 (1997).
There has been a long felt need for technology sufficient to allow for the advance warning of an impending asthma episode which would permit a patient either to immediately begin medication or to seek medical intervention. A surrogate for an indication of asthma treatment is the presence of nitric oxide in the human breath. However, an instrument capable of providing such warning would require the capability of measuring levels of nitric oxide of about 100 ppB with a resolution of less than about 10 ppB. Such an instrument must further detect these low levels of nitric oxide in the presence of other constituents of human breath such as water vapor and carbon dioxide. In addition, any such instrument should fee simple to use, maintain, and calibrate, thus making it useable in hospital or in home health care settings.
Using Lasers to Measure Component Concentrations in a Gas Sample
Spectroscopy has been used to determine the concentration of a component gas in a given sample for many years. Initially, spectroscopy was conducted using an infrared lamp as an energy source, and passing the light through a sample. The absorption, and thus the concentration of a component within the sample, is measured by normalizing the light energy remaining after passing through the sample with the light energy that entered the sample. Unfortunately, because of the difficulty in controlling the output of the source, and the need to have the light travel as great a distance as possible within a sample to maximize absorption, this process is incapable of easily measuring the concentration of trace components making up less than 1 ppm of a gas sample within a short time interval. See, for example, U.S. Pat. No. 3,792,272.
Over the last thirty years techniques have been developed that allow the measurement of component concentrations within a gas sample. One method, described in “Spin Flip Raman Laser and Infrared Spectroscopy,” Phys. Rev. Lett. 25:8-11 (1970), incorporated herein by reference, passes a laser beam from a tunable radiation source, the beam frequency corresponding to a fundamental absorption peak frequency (also known in the art as a vibrational-rotational peak frequency) of the component being measured, through a test sample to obtain the trace component concentration. The measurement is typically made by first splitting the beam into two parts, a first beam and a second beam, using a beam splitter having known beam splitting properties. The first beam is directed to a first detector where its power is measured. The power of the first beam is used in conjunction with the known properties of the beam splitter to determine the power of the second beam incident on the test sample. The second beam is passed through the test sample, where it is partially absorbed by the component in the test sample, resulting in an attenuation of the second beam's total energy. Upon emerging from the test sample, the energy of the second beam is measured by a second detector. The output of the second defector, therefore, contains the natural variations and fluctuations of the power of the laser beam diminished slightly due to attenuation from absorption in the test sample. The energy absorbed by the component in the test sample is derived from the difference between the output of the first and second detectors. The component concentration within the test sample is obtained by comparing the above absorption measurement with the absorption measurement obtained from a sample having a known component concentration.
The sensitivity afforded by this technique is, however, limited. When measuring small absorption amounts, i.e., the test sample has only trace amounts of the gas being measured, the power of the beam, both incident upon and emerging from the test sample, is very large in comparison to the amount of energy absorbed by the trace component. Therefore, the error margins present in the power measurements will have a larger effect on the calculation of the relatively small absorption amount. By way of example, this technique may be likened to determining the weight of the captain of an oil super tanker by weighing the tanker with and without the captain on the ship and subtracting the latter measurement from the former.
The sensitivity of the preceding technique may be enhanced by placing test sample and the laser within an optical cavity. The optical cavity is formed by two very highly reflective mirrors, each having a reflectivity of approximately between 99.95% and 99.99%. Within the optical cavity, the optical power passing through the test sample at any given time is greatly increased due to the beam reflecting back and forth between the mirrored surfaces. A small but measurable amount of energy from the beam passes through the mirror as “leakage”. This leakage is used to accurately determine the amount of energy circulating in the cavity according to well known principles in the art. To determine the absorption, and therefore the concentration, of a component within a test sample using an optical cavity, the leakage from the cavity in the absence of the test sample is measured, yielding the energy of the beam incident on the test sample, and subtracting from that the leakage measured in the presence of the test sample. However, even with the use of the optical cavity, this technique is limited because it measures absorption indirectly by taking the difference between the laser energy before it enters the sample and the laser energy after it passes through the sample,
Calorimetric Detection:
A second method, described in “Nitric Oxide Air Pollution: Detection by Optoacoustic Spectroscopy” Science 173:45-47 (1971), incorporated herein by reference, greatly increases the sensitivity of absorption measurements through the use of single-pass optoacoustic spectroscopy. In this method, an acoustic microphone is placed in a gas cell containing a test sample having an unknown concentration of a component gas. A pulsed or chopped tunable laser beam is passed through the gas ceil and the energy absorbed by the component gas is directly measured using the acoustic microphone. If the test sample contains non-absorbing gases or the frequency of the beam does not correspond to a fundamental absorption peak of any gas within the test sample, including the component gas, the beam exits the gas cell unattenuated. However, if the test sample contains a component gas that is absorbing and the beam frequency corresponds to a fundamental absorption peak frequency of the component gas, energy from the beam is absorbed by the component gas. This energy absorption causes slight heating within the test sample that occurs at regular and periodic intervals because the beam is pulsed and no absorption or heating occurs between pulses. The periodic heating of the test sample causes pressure fluctuations to be generated and propagated within the gas cell. These pressure fluctuations are sound waves and are detected by the microphone within the gas cell. The concentration of the component gas is determined by normalizing the acoustic energy, as measured by the microphone, with the energy of the beam incident on the gas cell, and comparing the result with a similar measurement using a sample having a known component concentration. This method provides a direct measurement of energy absorption in the cell allowing the measurement of component concentrations making up as little as 1 ppB of the test sample.
The sensitivity of optoacoustic spectroscopy may be further enhanced by placing the gas cell containing the test sample in an optical cavity such as the one previously described. The pulsed beam is directed into the optical cavity in the manner previously described, and if the test sample contains a component gas that is absorbing and the beam frequency corresponds to a fundamental absorption peak frequency of the component gas, then the component gas absorbs energy from the beam. As in the previous method, the microphone is used to measure the acoustic energy. The energy of the beam incident on the gas cell is determined by measuring the leakage from the cavity in the absence of the test sample. The concentration of the component gas may then be obtained in the manner previously described. This variation of optoacoustic spectroscopy allows the detection of trace components that make up as little as 0.1 ppB (or 100 parts per trillion) of the test sample.
The high sensitivity spectroscopy methods described above, however, have been used chiefly for studying contaminant concentrations in the atmosphere and have found few applications in other fields.
Measuring Component Concentrations in Human Breath
Up until the last few years, one of the only applications of spectroscopy as applied to measuring the component concentration of human breath has been testing for ethyl alcohol, as described in U.S. Pat. No. 3,792,272. This application has found practically universal use by law enforcement agencies everywhere, but other uses and users are virtually non-existent. The technique disclosed in U.S. Pat. No. 3,792,272 passes frequency-modulated radiation in the infrared range, either from a laser or a collimated lamp, through a sample in order to determine the blood-alcohol content of an individual by measuring ethyl alcohol in the individual's breath. An individual being tested breathes info a heated collection chamber which keeps the breath from condensing and allows for a more accurate measurement. The collection chamber also contains two mirrors to reflect the light source back and forth and provide it with a longer path length in the collection chamber. The amount of radiation energy exiting the collection chamber is measured and then normalized using the radiation energy that entered the collection chamber, yielding the amount of energy absorbed by the ethyl alcohol in the sample. However, for the same reasons previously described, this method is not capable of detecting minute component concentrations within a breath sample.
U.S. Pat. No. 4,314,564 discloses improvements to the alcohol breath test described above. These improvements, however, relate only to eliminating the need to heat the collection chamber. This improved device eliminates the need to heat the chamber by accounting for the relative humidify within the gas chamber and the ambient atmospheric humidity. However, in all other regards, the improved alcohol breath test is the same as the previous one and is incapable of measuring minute component concentrations.
More recently, human breath has been examined for the presence of isotopes or isotopically labeled molecules, as disclosed in U.S. Pat. Nos. 5,543,621 and 5,961,470, respectively. Both of these human breath tests, however, are not capable of detecting trace component concentrations in exhaled human breath because they rely on the same general techniques discussed used in the detection of alcohol. Additionally, U.S. Pat. No. 5,543,621 only measures the ratio of the concentration of the isotopically substituted component to the concentration of the more common form of the element; it does not independently measure the actual concentration of the isotope.
Another recent use of spectroscopy to measure gaseous components of human breath utilizes complex mathematical approximation methods to arrive at the concentration of the gaseous components, such as is disclosed in U.S. Pat. No. 5,807,750. This method passes multi-spectral collimated light through a gas sample and detects the power of the emerging light using an array of detectors, with each detector in the array set to detect the emerging light at a single frequency. The signal obtained from the defector array is passed to a computer system which performs the complex mathematical calculations to arrive at the concentration of the component gases. The complex mathematical calculations are based on experimentally derived algorithms for several species of gases, including the component of interest being measured and many possible interfering species. The experimentally derived algorithms form a matrix which the computer system uses in an iterative process to determine, from the output of the detector array, a component absorption approximation free of absorption from interfering species. The absorption approximation is thereafter used to determine the concentration of the component of interest. This method, however, may not be ideal for use under all circumstances because the complex mathematical analysis requires the presence of a computer capable of performing such analysis and the creation of experimentally derived algorithms for each component measured and all possible interfering species.